Skip to main content
  1. Knowledge base
  2. Findings
Check: email.dane.unsigned_zone

Fix TLSA Record Not Anchored by a DNSSEC Signed Zone

A TLSA record protects inbound mail only when it sits in a DNSSEC signed zone whose chain of trust validates from the parent. If the zone is unsigned, the DS record is missing, or the parameters are ones senders ignore, DANE-aware servers skip the record and fall back to unauthenticated TLS. Fix it by signing the zone, publishing the DS, and using usage 3 or 2.

What is a TLSA record, and why does it need DNSSEC?

A TLSA record is a DNS resource record that binds a mail server's TLS certificate or public key to its hostname, so a sending server can verify it without trusting any certificate authority. This is DANE for SMTP, described in RFC 7672, section 2.

The binding is only as trustworthy as the DNS answer that carries it. An attacker who can forge DNS responses could also forge a TLSA record that matches their own certificate. DNSSEC closes that gap, so RFC 7672 has senders use TLSA data only when the response validates as secure. An unsigned TLSA record is treated as if it did not exist.

What causes this finding?

The scanner raises it when TLSA records are published for a domain's mail hosts but the protection they advertise cannot work. Typical causes:

  • The zone holding the TLSA record is not signed at all.
  • The zone is signed but no DS record was added at the parent, so validators see an insecure delegation.
  • A DS record exists but does not match the current key, which makes the zone bogus rather than merely insecure.
  • The TLSA record uses parameters a conforming sender will not honour, such as certificate usage 0 or 1.
  • No successor record is staged ahead of a certificate renewal.

The mail host's own zone matters here, not only the zone of the domain in the recipient address. The TLSA record lives under the MX hostname, and that name's zone must validate.

_25._tcp.mx1.example.com. 3600 IN TLSA 3 1 1 (
    8cb0fc6c527506a053f4f14c8464bebbd6dede2738d1 1d2e0d5a7e5b0f4a2c91 )

The three numbers are usage, selector and matching type. Usage 3 is DANE-EE, selector 1 is the public key, and matching type 1 is SHA-256. RFC 7672 recommends usage 3 or 2 for SMTP.

Why is it rated high?

The failure is silent in one direction and loud in the other.

If the zone is unsigned, nothing breaks. Mail flows, the TLSA record looks correct, and the operator believes STARTTLS downgrade attacks are covered. But senders ignore the record, so an on-path attacker can still strip STARTTLS or present a forged certificate. The article on the STARTTLS downgrade attack that a missing TLSA record allows shows the mechanics.

If the zone is signed but broken, for example a DS that no longer matches, validating resolvers return a failure. Senders that validate DNSSEC may then defer or bounce mail. That costs you an outage, and an auditor reviewing your mail transport controls will count the unprotected case as a false assurance.

How does the scanner detect and rate it?

The check queries the TLSA records for each MX host on port 25 and evaluates three things for each:

  1. Anchoring. Whether the TLSA answer validates as secure, which requires a signed zone and a matching DS at the parent.
  2. Parameters. Whether at least one record uses a usage, selector and matching type that SMTP senders honour.
  3. Renewal safety. Whether a second record is staged alongside the current one.

The finding carries the MX hostname, the TLSA owner name, the parameter triples found, and the DNSSEC validation status observed. Severity is high when TLSA records are published but not anchored. Exact evidence fields and resolver behaviour can differ between scanner versions, so read the evidence on your own report instead of assuming a layout.

How do I fix a TLSA record that is not in a DNSSEC signed zone?

Work through these steps in order.

  1. Sign the zone. Enable DNSSEC at your DNS provider or signer, using a current algorithm your provider supports.
  2. Publish the DS record at the parent. Take the DS data from your signer and add it at your registrar. Without this step the zone stays insecure. The guide on fixing a missing DS record at the parent zone covers it in detail.
  3. Use sender-honoured parameters. Prefer 3 1 1 or 2 1 1. Avoid usages 0 and 1.
  4. Stage the successor before renewal. Publish a record for the new key alongside the current one, wait for the old TTL to expire, then deploy the new certificate. Reusing the same key across renewals avoids the problem for selector 1. The outage pattern is described in what happens when a certificate is rotated and the TLSA record is forgotten.

To compute the record data from the live certificate:

openssl s_client -starttls smtp -connect mx1.example.com:25 -servername mx1.example.com </dev/null 2>/dev/null \
  | openssl x509 -pubkey -noout \
  | openssl pkey -pubin -outform DER \
  | openssl dgst -sha256 -binary | xxd -p -c 64

Add this hash to the 3 1 1 record, and keep both old and new hashes during a rotation.

How do I confirm the fix on re-scan?

First verify what a validating resolver sees. The ad flag must be set on the TLSA answer:

dig +dnssec TLSA _25._tcp.mx1.example.com @1.1.1.1

Look for ad in the flags line and an RRSIG alongside the TLSA record. Then check that the zone has a DS at the parent:

dig +short DS example.com

Re-run the scan after the DS and TLSA TTLs have expired. The finding clears when the answer validates as secure, a usable parameter set is present, and no renewal gap is detected.

Common questions

Can I publish TLSA records without DNSSEC?

You can, but it does nothing useful. RFC 7672 has senders use TLSA data only when DNSSEC validation says the answer is secure. An unsigned record is ignored, so you get no downgrade protection while appearing to have it. Sign the zone first, then publish TLSA.

Which TLSA parameters should I use for SMTP?

Use 3 1 1 for DANE-EE with a SHA-256 hash of the public key, or 2 1 1 to anchor on a CA certificate in the chain. RFC 7672 tells operators not to publish usages 0 and 1 for SMTP, because senders do not treat them as authenticating records.

Will a broken DNSSEC chain stop my mail?

It can. If the zone is signed but the DS no longer matches, validating resolvers report failure rather than an unsigned answer. Senders that validate may then defer delivery. Treat DS changes and key rollovers as mail-affecting operations, and verify with dig right after each one.

Key takeaways

  • A TLSA record only counts when its zone validates under DNSSEC from the parent's DS record.
  • Unsigned means silently ignored. A broken chain can mean deferred mail.
  • Use 3 1 1 or 2 1 1, and stage the successor record before each certificate renewal.
  • Confirm with dig +dnssec and look for the ad flag.
  • SetEnforce scans each MX host's TLSA records and checks that they are signed, usable, and renewal-safe.

Check whether your own domain gets this right. Scanning is free and needs no account.

Scan your domain

All articles